Three iterations of Tween Day were completed by the research team with three unique tween samples. The locations chosen were Juicy Couture Jewellery a university campus in the city (hereafter “University”), a faith-based ministry in a culturally diverse urban neighborhood (“Ministry”), and an elementary school in a middle-class suburb twenty miles from the city (“School”). In total, thirty-four tweens participated: sixteen at University (ten females and six males; average age 11 years), five at the Ministry (two females and three males; average age 12 years), and thirteen at the School (seven females and six males; average age 11 years). The University tweens were Caucasian; the Ministry tweens were African American; and the School tweens were a mix of Caucasian, African American, Native American, and Asian/Pacific Islander. Although the participants were recruited conveniendy at each location, the combined sample of thirty-four tweens represented a diverse range of socioeconomic and ethnic groups, one roughly equivalent to the study’s geographic region, which was broadly situated in a large urban-suburban area with a population in excess of 3 million.

By systematically collecting data from three sites using identical instruments, protocols, and time frames, the research team was able to perform a thorough, cross-site analysis of the qualitative data. All data collected were Juicy Couture Necklaces rigorously checked for validity. To ensure trustworthiness, we implemented several measures as recommended by Yvonne Lincoln and Egon Cuba [49]. Dependability (or reliability) was ensured through:
Consistent note-taking and the use of multiple researchers
Using multiple, triangulated methods
Comparing emergent themes with findings from related studies
Audiotaping and transcribing interviews
Employing intercoder checks
Analyzing the data for incidents of observer effect
We addressed different forms of validity as follows:
Face validity: asked whether observations fit an expected or plausible frame of reference Criterion/internal validity: (1) pretesting instruments, (2) peer debriefing, and (3) participant verification (i.e., member checks)Construct validity: examined data with respect to the various theoretical constructs represented in the literature and pertinent to different aspects of the study (e.g., information grounds)

The qualitative data set was analyzed using a set of coding techniques, memos, and research team conversations. Major themes emerged from the data, which, at a high level, guided the analysis and permitted the team to break an enormous data set into manageable portions. These themes (e.g., information grounds, social types, affect, and information needs) or first-level codes were mapped onto the instruments’ questions to create thematic “sets” that could be analyzed using iterative pattern-coding techniques. The researchers utilized the coding and querying features of Atlas Ti 5.0 (Knowledge Workbench, Inc.) to parse the data into these thematic sets. Repeated reading of the interview transcripts, memos, and group discussions allowed the researchers to identify patterns (second-level codes) of behavior within these sets. Some patterns appeared in multiple themes, which permitted thematic bridging. Select questions that provided particularly rich data were further analyzed independent of the themes, using third-level coding schemes specific to those smaller analytic sets. The complete methodology (including instruments and techniques) is explained in detail in Meyers, Fisher, and Marcoux [50]. The resulting data amassed from fourteen focus groups and twenty-five interviews were analyzed using the aforementioned frameworks and principles, yielding unique insights into the information worlds of tweens and its social, affective, and cognitive dimensions.